https://github.com/valenradovich/talk-to-link
Talk to as many links as you want. Get a summary, ask specific questions, explore the depths of the information. Your imagination is the limit.
https://github.com/valenradovich/talk-to-link
langchain llm openai rag streamlit
Last synced: 3 months ago
JSON representation
Talk to as many links as you want. Get a summary, ask specific questions, explore the depths of the information. Your imagination is the limit.
- Host: GitHub
- URL: https://github.com/valenradovich/talk-to-link
- Owner: valenradovich
- License: mit
- Created: 2023-11-24T15:59:26.000Z (over 2 years ago)
- Default Branch: main
- Last Pushed: 2024-01-09T00:44:24.000Z (over 2 years ago)
- Last Synced: 2025-01-20T01:26:20.457Z (over 1 year ago)
- Topics: langchain, llm, openai, rag, streamlit
- Language: Jupyter Notebook
- Homepage: https://talk-to-links.streamlit.app
- Size: 455 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Talk to Links 💬
Talk to as many links as you want. Get a summary, ask specific questions, explore the depths of the information. Your imagination is the limit.
## Why?
- Too many websites of your interest to read?
- Too many news to follow?
- Too many Medium posts saved and not enough time to read them all?
## How to use
1. Go to [Talk to Links](https://talk-to-links.streamlit.app) website.
2. Enter your [OpenAI API](https://platform.openai.com/account/api-keys) key.
3. Enter the URLs you want to process.
4. Done! Just chat as you do regularly to Chat GPT.
## Demo
In this demo i'm doing an example of how someone can use the app, doing specific questions or just asking for a summary
[demo_talk_to_link.mp4](https://github.com/valenradovich/talk-to-link/assets/86897297/501f1862-2477-489d-8d9e-b2db936f776b.mp4)
## Technologies
It uses [OpenAI's API](https://openai.com/), [LangChain](https://www.langchain.com) and [FAISS](https://github.com/facebookresearch/faiss).